Machine learning for stock market réalisé par un freelance en Tunisie | Tunisie Freelance
Machine learning for stock market
data-science
Rôle: Developer
LSTM for 24-Hour EUR/USD Price Forecasting with 100% accuracy in prediction trends and 90+\~ accuracy in predicting movement as showed in the pictures
Excited to share a project I’ve been working on recently: building a deep learning model to predict 24-hour high and low price movements for EUR/USD using MetaTrader 5 data.
Using over 4,000 historical H1 bars, I built and trained two LSTM models to forecast the next 24 hourly highs and lows, leveraging a 3-day lookback window. The model architecture uses stacked LSTM layers and is trained to minimize MSE for both high and low price targets.
PythonMachine Learning
Pièces jointes
Z
Software Engineering Student | AI & Machine Learning Specialist